Search

This study examined an area of woodland that was recovering from severe fire in Royal National Park (NSW, Australia). A non-destructive method of field sampling is required for vulnerable recovering vegetation and therefore classification of digital photographs using linguistic terms was trialled. The linguistic data for three vegetation strata (canopy, shrub and ground) were converted to crisp scores and compared with vegetation index data derived from remotely sensed imagery. All possible subset regression was used to test the proposition that the combined vegetation scores (independent variables) would explain the values of NDVI (Normalized Difference Vegetation Index) and NDMI (Normalized Difference Moisture Index). Vegetation scores for the three strata were also combined using simplified weighting ratios to assess broad relationships between the indices and field data. The combined vegetation scores explained ∼60% of the variation in the vegetation index data and inclusion of variables representing multiple strata explained more of the variation than any single variable. The precise value of the weights used to combine the layers did not affect the strength of the association. A simple ratio is proposed that may be useful to estimate woodland parameters under similar conditions, by inversion of the relationship with vegetation index data. [ABSTRACT FROM AUTHOR]

Abstract: This study analyzed the geographical patterns of heat-related mortality among the population aged 65 and over within the metropolitan area of Sydney, Australia between 1993 and 2004, and evaluated the role of some physical and socio-demographic risk factors associated with it. The effect of temperature on all-cause mortality during unusually hot days was investigated using spatial analytic techniques, such as cluster analysis and spatial regression analysis. Generalized Linear Models (GLMs) were used to investigate the role of daily average temperature, ozone (O3) and particulate matter of diameter less than 10μm (PM10) at the regions that showed a significant increase in mortality on unusually hot days. Spatial variation in mortality on unusually hot days was observed among the population 65 and over. Elderly people living within 5–20km south-west and west of the Sydney Central Business District (CBD) were found to be more vulnerable. However, analysis using GLMs showed temperature to be a significant modifier of daily mortality in the region to the south-west of the CBD only. O3 and PM10 were found to be non-significant factors in the regions where air pollutants were studied. Socio-economic status and the proportion of vegetation or developed land in each Statistical Local Area (SLA) were also not a significant factor explaining the increased mortality. A combination of social and environmental factors may be at play. Our results suggest an effect of temperature on mortality of the elderly population in Sydney Statistical Division at the SLA level. More spatially-based research would be beneficial once climate datasets with improved spatial coverage become available. [Copyright &y& Elsevier]